Btrfs: fix how we do space reservation for truncate

The ceph guys keep running into problems where we have space reserved in our
orphan block rsv when freeing it up.  This is because they tend to do snapshots
alot, so their truncates tend to use a bunch of space, so when we go to do
things like update the inode we have to steal reservation space in order to make
the reservation happen.  This happens because truncate can use as much space as
it freaking feels like, but we still have to hold space for removing the orphan
item and updating the inode, which will definitely always happen.  So in order
to fix this we need to split all of the reservation stuf up.  So with this patch
we have

1) The orphan block reserve which only holds the space for deleting our orphan
item when everything is over.

2) The truncate block reserve which gets allocated and used specifically for the
space that the truncate will use on a per truncate basis.

3) The transaction will always have 1 item's worth of data reserved so we can
update the inode normally.

Hopefully this will make the ceph problem go away.  Thanks,

Signed-off-by: Josef Bacik <josef@redhat.com>

1 files changed, 36 insertions, 10 deletions

diff --git a/fs/btrfs/extent-tree.c b/fs/btrfs/extent-tree.c
index ca599654ce19..a2ca561c70f0 100644
--- a/fs/btrfs/extent-tree.c
+++ b/fs/btrfs/extent-tree.c
@@ -3980,6 +3980,37 @@ static u64 calc_trans_metadata_size(struct btrfs_root *root, int num_items)
 		3 * num_items;
 }
 
+int btrfs_truncate_reserve_metadata(struct btrfs_trans_handle *trans,
+				    struct btrfs_root *root,
+				    struct btrfs_block_rsv *rsv)
+{
+	struct btrfs_block_rsv *trans_rsv = &root->fs_info->trans_block_rsv;
+	u64 num_bytes;
+	int ret;
+
+	/*
+	 * Truncate should be freeing data, but give us 2 items just in case it
+	 * needs to use some space.  We may want to be smarter about this in the
+	 * future.
+	 */
+	num_bytes = calc_trans_metadata_size(root, 2);
+
+	/* We already have enough bytes, just return */
+	if (rsv->reserved >= num_bytes)
+		return 0;
+
+	num_bytes -= rsv->reserved;
+
+	/*
+	 * You should have reserved enough space before hand to do this, so this
+	 * should not fail.
+	 */
+	ret = block_rsv_migrate_bytes(trans_rsv, rsv, num_bytes);
+	BUG_ON(ret);
+
+	return 0;
+}
+
 int btrfs_trans_reserve_metadata(struct btrfs_trans_handle *trans,
 				 struct btrfs_root *root,
 				 int num_items)
@@ -4020,23 +4051,18 @@ int btrfs_orphan_reserve_metadata(struct btrfs_trans_handle *trans,
 	struct btrfs_block_rsv *dst_rsv = root->orphan_block_rsv;
 
 	/*
-	 * one for deleting orphan item, one for updating inode and
-	 * two for calling btrfs_truncate_inode_items.
-	 *
-	 * btrfs_truncate_inode_items is a delete operation, it frees
-	 * more space than it uses in most cases. So two units of
-	 * metadata space should be enough for calling it many times.
-	 * If all of the metadata space is used, we can commit
-	 * transaction and use space it freed.
+	 * We need to hold space in order to delete our orphan item once we've
+	 * added it, so this takes the reservation so we can release it later
+	 * when we are truly done with the orphan item.
 	 */
-	u64 num_bytes = calc_trans_metadata_size(root, 4);
+	u64 num_bytes = calc_trans_metadata_size(root, 1);
 	return block_rsv_migrate_bytes(src_rsv, dst_rsv, num_bytes);
 }
 
 void btrfs_orphan_release_metadata(struct inode *inode)
 {
 	struct btrfs_root *root = BTRFS_I(inode)->root;
-	u64 num_bytes = calc_trans_metadata_size(root, 4);
+	u64 num_bytes = calc_trans_metadata_size(root, 1);
 	btrfs_block_rsv_release(root, root->orphan_block_rsv, num_bytes);
 }